This content has been marked as final. Show 6 replies
Which database version are you using? OS?
Hi;1 person found this helpful
Thanks Helios. I'm aware of the documentation/link that you sent me.
I'm looking for an explanation of what might've happened in the scenario that I mentioned above and what would be a better option (Would KILL_SESSION always work as opposed to CANCEL_SQL?, Are there other options I could use to make sure that a User query is killed before it delays a process of a higher priority?)
I guess that would come down to a DBA's experience and not a textbook answer.
11.2 is pretty stable. If you have checked MOS note ID 1340184.1, almost all "Recommended Patches for Managing Runaway Queries with Oracle Database Resource Manager" are fixed on 11.0.7. Also, you have the view v$rsrcmgrmetric_history to monitor usage of CPU and waits.
Without a test case, we are assuming that it is configured correctly. That is something you need to double-check to be sure that your plan is active and the query is in the right group.
For this, MOS ID 1358709.1 is a good starting point.
I've seen heavy loaded servers who take some extra time to do the switch. There you need to do a closer monitoring of the system (waits) and the session being switched to figure out what is happening, and see if you can co-relate this.
gotcha! Thank you!
I'll look into it!